Specific Aims: The overarching goal of this proposal is the development of a reliable strategy for differentiation of central pain predominant from peripheral pain predominant knee osteoarthritis (KOA), chronic low back pain (CLBP) and painful diabetic neuropathy (DPN) patients using clinical features, experimental pain testing and magnetic resonance spectroscopy (MRS). We will study both excitatory (glutamate + glutamine (Glx) and N-acetylasparatylglutamate (NAAG)) as well as inhibitory (gamma-Aminobutyric Acid (GABA)) neurotransmitters using established and novel MRS techniques. Voxels for MRS will be placed in brain regions implicated in pain processing including the anterior insula, posterior insula, anterior cingulate cortex, and thalamus. The following specific aims are proposed: 1. Demonstrate that distinct subgroups of KOA, CLBP and DPN patients exhibit a symptom profile and experimental pain testing parameters suggestive of significant CNS contribution to their pain (i.e. significant central pain component). 2. Using conventional MRS techniques, determine if central pain predominant chronic pain states have elevated levels of Glx, a major excitatory component in the central nervous system. 3. Using novel MRS techniques, determine if central pain predominant chronic pain states have reduced levels of GABA which is a major inhibitory neurotransmitter and lower levels of NAAG, an antagonist to glutamate. Study Design & Methods: The study will use a cross sectional design comparing 70 subjects with KOA, 70 subjects with CLBP, 30 subjects with DPN and 30 pain free controls. Subjects will be recruited over a four year period; patient subjects will come from the Ann Arbor VA Pain Clinic. Subjects will undergo clinical pain questionnaires, clinical pain experiment characterization and established and novel (MEGA-PRESS) MR spectroscopy techniques to assess for central versus peripheral pain predominance. A logistic regression classifier based on pain measures will be developed using logistic discrimination to determine peripheral versus central predominance of pain symptoms. A cross-sectional regression analysis will be used to assess the association of Glx and GABA with subject group (central pain predominant, peripheral pain predominant, healthy controls) and other explanatory variables. Potential confounders such as pain and psychiatric medications will be controlled for using logistic regression adjustments/stratification measures. Significance: Chronic pain syndromes are common in the Veteran population and are often refractory to available treatments and interventions. It is estimated that 30-40% of pain syndromes thought generally to be centrally mediated have an important central component; this is an important distinction as treatment for central and peripheral pain predominant entities are distinctictly different. Evaluation of both the neuroinhibitory (GABA) and neuroexcitatory (Glx & NAAG) pathways provides promising potential both in terms of improving our understanding of the mechanisms of chronic pain as well as opportunities for effective, individual-based treatments.